The rise of cryptographic agents focused on AI follows a familiar trajectory which reflects the initial boom, the bust and the resurgence of ICO projects. Just as the first blockchain companies prospered on media threshing before ripening in sustainable ecosystems, the current wave of AI agent projects undergoes rapid market changes.
A new HTX Ventures and HTX Research report indicates that investors are increasingly cautious because competition in the sector is intensifying, liquidity dispersions and many projects have difficulty defining clear use cases. However, while the sector goes beyond its speculative phase, Cryptographic agents focused on AI should change sustainable commercial models underlying by real utility.
To dive deeper into the evolution of cryptographic agents and the future of blockchain innovation motivated by AI, download the full HTX report here.
From media threshing to reality: the evolution of cryptographic agents
The initial wave of projects of cryptographic agents in 2024 was motivated by the blind enthusiasm for AI projects. After the impact of a bitcoin donation of $ 50,000 from Marc Andreessen in October 2024 and the success of Token Launchpads earlier in the year, many AI agent projects entered the Q1 space of 2024 and liquidity quickly diluted by T1 of 2025. As for any emerging sector, Hye in the early state has not always translated the long -term viability, and a period Cooling and a long -term violation period, and a cooling period, and a period of long -term viability, and a period of cooling, and a long -term viability period, and a period of cooling and a period of long -term viability, and a period of cooling and a period of long -term viability, and a period of cooling and a period of long -term viability, and a period of cooling in crypto.
The market segment now enters a more mature phase, and the emphasis is put from speculative excitement to revenue generation and product performance. The winners of this scalable landscape will be those who can generate stable income, cover the management costs of AI models and provide tangible value to users and investors.
AI agent applications emphasize the implementation and marketing of the real world of this technology, in particular in fields such as automated trade, asset management, market analysis and cross -interaction. This approach aligns with multi-agent systems and Defai (decentralized + AI finance) like Hey Anon, Griffain and Chaingpt.
Recent research highlights the advantages of multi-agent systems (Mas) in portfolio management, especially in cryptocurrency investments. Projects such as Griffain, Neur and Buzz have already demonstrated how AI can help users interact with DEFI protocols and make informed decisions. Unlike single agent AI models, multi-agent systems use collaboration between specialized agents to improve market analysis and execution. These agents operate in teams, such as data analysts, risk assessors and trading execution units, each formed to manage specific tasks.
MAS executives also introduce inter-agent communication mechanisms, where agents within the same team refine predictions by collective learning, reducing errors in the analysis of market trends. The next Defai phase will likely imply a decentralized governance models, where multi-agent systems participate in the management of the protocol, the optimization of the treasury and the application of onchain compliance.
To dive deeper into the evolution of cryptographic agents and the future of blockchain innovation motivated by AI, download the full HTX report here.
Deepseek-R1: a breakthrough in the training of AI agents
A breakthrough in AI agents technology has arrived with Deepseek-R1, an innovation that questions traditional AI training methods. Unlike the previous models, which were based on a supervised fine setting (SFT) followed by the learning of strengthening (RL), Deepseek-R1 adopts a different approach, completely optimizing by learning strengthening without initial supervised phase. This change has led to remarkable improvements in the capacities of reasoning and adaptability, paving the way for cryptographic agents more sophisticated by AI.
To understand this paradigm shift, consider two different approaches to learning. In the traditional SFT and RL model, a student first studies from a workbook, practicing problems with the defined answers (SFT), then receives tutoring to refine their understanding (RL). On the other hand, with the Deepseek-R1 model (pure learning by strengthening), the student is thrown directly into an exam and learns by trials and errors. This approach allows the student to improve dynamically according to comments rather than relying on predefined responses.
Taking advantage of the Pure RL model of Deepseek-R1, AI agents learn by trials and errors under real conditions, dynamically adjusting their strategies according to immediate feedback.
This method allows greater adaptability, which makes it particularly useful for multi-agent AI systems in DEFI, where market fluctuations in real time require agents to take autonomous and data decisions. For example, agents supplied by AI can monitor liquidity pools, detect arbitration opportunities and optimize asset allowances according to real -time market conditions. These agents quickly adapt to market fluctuations, ensuring a more effective deployment of capital.
Launched at the end of November 2024, Idegen was the first agent of the Crypto AI built on Deepseek R1. This integration of the DEEPSEEK R1 model underlines how Crypto AI agents can inherit these improved reasoning capacities, in competition with other models of AI established at a cost fraction.
This transition to a multi-agent AI fueled by RL in DEFI automation underlines why the models of closed source (such as OPENAI GPT systems) become an unsustainable expenditure. Work flows often require treatment of more than 10,000 tokens per transaction, closed AI models impose significant calculation costs, limiting scalability. On the other hand, the RL Open Source models as Deepseek-R1 allow a decentralized and profitable AI development adapted to DEFI applications.
The future of AI agents in web3
The key to longevity in this sector lies in continuous innovation, adaptability and profitability. Open source models like Deepseek-R1 reduce barriers to entry, allowing native blockchain startups to develop specialized AI solutions. Meanwhile, the progress of the Defai and Multi-Aggen systems will lead to long-term integration between AI and decentralized finance.
The point to remember is clear: projects must prove their value beyond the media threw. Those who develop sustainable economic models and exploit advances in advanced AI will define the future of intelligent blockchain ecosystems. The ICO era of crypto agents is evolving, and the next wave of winners will be the one that can transform innovation into long -term viability.
To dive deeper into the evolution of cryptographic agents and the future of blockchain innovation motivated by AI, download the full HTX report here.
Non-liability clause. Cointtelegraph does not approve of any content or product on this page. Although we aim to provide you with all the important information that we could obtain in this sponsored article, readers should do their own research before taking measures related to the company and to win full responsibility for their decisions, and this article cannot be considered as an investment advice.
This article is for general information purposes and is not intended to be and must not be considered as legal or investment advice. The points of view, the thoughts and opinions expressed here are the only of the author and do not reflect or do not necessarily represent the opinions and opinions of Cointellegraph.
Cointtelegraph does not approve of the content of this article or any product mentioned here. Readers must do their own research before taking measures related to any product or company mentioned and assume full responsibility for their decisions.
